If you haven't studied semiotics(or even if you have!), you may be confused by an important distinction we will be making between the semiotics of an image and its grammar. Here is a relatively simple way to understand that distinction.
Semiotics was formulated by the linguist Ferdinand de Saussure (let's leave aside for now its earlier formulation by C.S. Pierce). Saussure wanted to understand how meanings (signifieds) map to speech sounds (signifiers - e.g. words or lexemes). What struck him is that the signifiers are arbitrarily related to what they signify, and yet the signified is effectively impressed on the signifier (the material entity - an acoustic event or marks on a page). We experience the signifier as if it just were the signified. Indeed, it is virtually impossible not to hear the "meaning" of lexemes in your native language.
Saussure also noted something else of importance. Signifiers tend to form closed systems, i.e. there are only so many elements in a set of possible signifiers - a deck of cards, allowed sounds in a particular language. Thus the value of a particular signifier can often be defined by its position in a system - the ten of clubs in a standard deck of cards.
Given this framework, an obvious question to ask is whether the signifieds are also part of closed systems. Unfortunately, many people who pursue this question fail to understand the basic semiotic paradigm and therefore conflate two levels of analysis. The fact that the relationship between signifiers and signifieds in natural speech is arbitrary does not mean this relationship will be arbitrary in other semiotic systems. Futhermore, the fact that signifiers are part of a closed system does not mean that the signifieds must also belong to a structured system of meanings.
To understand the importance of this point, consider the following example. In the simple system of signifiers in the table on the left, I selected five vowel sounds from the phonological feature space shown on the right and placed each one between two consonants - "p" and "n." The signifieds for these five lexemes are the core entries each one generated in the online Marion Webster dictionary.
The first column in the table on the left illustrates the principle of positional variation for signifiers. The five possible signifiers in this simple system are differentiated purely on the basis of contrastive distinctions between the positions of vowels in the feature space on the right, e.g. pen / pin. The mapping of the signifieds to these five signifiers illustrates the principle of arbitrariness. There is no intrinsic or motivated reason other than purely contingent historical factors for any of the possible "meanings" to map to any of the signifiers. In fact, one of the possible signifiers - pon - has no signifieds. Moreover two of the signifiers are highly polysemic - i.e. map to many different entries in a lexicon - while others such as "pun" are less so.
If you look at the dictionary entries for the highly polysemic signifiers such as "pan," you will very quickly see that it is not possible to make generalizations about either the arbitrariness or the semantic structure of the signifieds. Consider, for example, just one of the entries for "pan":
1 a: a usually broad, shallow, and open container for domestic use (as for cooking ) b: any of various similar usually metal receptacles: as (1): the hollow part of the lock in a firelock or flintlock gun that receives the priming (2): either of the receptacles in a pair of scales (3): a round shallow usually metal container for separating metal (as gold) from waste by washing cBritish: toilet 3b; also: bowl 3b d : steel drum 2 a(1): a natural basin or depression in land (2): a similar artificial basin (as for evaporating brine) b: a drifting fragment of the flat thin ice that forms in bays or along the shore.
There is an obvious motivated connection for most of the meanings in 1 and 2, namely what psychologists call affordance , i.e. "a quality of an object, or an environment, that allows an individual to perform an action." Even the signified in 2b can be taken as a semantic extension of the notion of a broad, shallow container.
Yet, even though many of these signifieds for "pan" have a deep motivated connection, they do not, either with respect to each other or within their own frame of reference, demonstrate the kind of positional variation found in the feature space for vowels. In other words, cooking pans, gun pans, and mining pans do not form a closed system. Moreover the domain of cooking utensils can be carved up many different ways by different cultures or at different periods of time (or by different manufacturers). Thus the contrastive distinction "pot / pan" at the level of signifieds is very much not the same as the distinction at the level of signifiers.
Semiotics is thus concerned with the relationship between signifiers and signifieds, a relationship that couples two very different things. The lesson here is that social selection on signifiers and signifieds also operates differently. Selection on a closed system of signifiers requires a remapping over a finite set of elements; selection on signifieds is highly contingent and does not necessarily involve a remapping across an entire feature space.
Grammar
A grammar is an account of how we map within the domain of signifiers or signifieds from an underlying representation (UR) to a surface representation (SR). In the tree diagram below, the German phrase "Wirtschafts und Währungs union" is the SR and the tree structure is the UR which maps the semantic properties in the set below the phrase to the SR.
"Wirtschafts und Währungs Union" (economic and monetary union)
We can also construct a grammar of the signifiers from the semiotic example, that is a grammar of how vowels and consonants combine to form syllables. Below are two examples of such a grammar - one from a classical treatment of phonology and the other from a more recent neural network based account of phonology called Optimality Theoretic grammar.
What Kress is concerned with in Reading Images is the grammar of how signifiers in an image go together to make a representation. This level of analysis must be distinguished from any sort of semiotic "interpretation" we wish to give to the signifieds associated with the image.
MUTTOPIA FLEECE JOGGING SUIT: These fleece lined jackets from the makers of Muttluks Boots are really nice! Made of warm, soft fleece lined material, these jackets are sure to keep your pet warm this winter. Great for indoor and outdoor use! These jackets velcro along the dog's back to make it easy to put on. If the legs are a little too long, you can just tuck them up under the elastic bands. You'll want a suit made for yourself!
We are Proud of Our Dogs! Bluegrass Kennel and Game Farm: Jim & Eunice Mell. 22493 181st Avenue, Sebeka, MN 56477
unstructured analytic processs
symbolic attributive
demand / social / involvement / equality
naturalistic coding orientation - "just being themselves"
Dog
Function: noun Usage: often attributive Etymology: Middle English, from Old English docga Date: before 12th century 1 a:canid; especially: a highly variable domestic mammal (Canis familiaris) closely related to the gray wolf b: a male dog ; also: a male usually carnivorous mammal 2 a: a worthless or contemptible person b:fellow, chap<a lazy dog> <you lucky dog> 3 a: any of various usually simple mechanical devices for holding, gripping, or fastening that consist of a spike, bar, or hook b:andiron 4: uncharacteristic or affected stylishness or dignity <put on the dog> 5capitalized: either of the constellations Canis Major or Canis Minor 6plural:feet 7plural:ruin<going to the dogs> 8: one inferior of its kind <the movie was a dog>: as a: an investment not worth its price b: an undesirable piece of merchandise 9: an unattractive person; especially: an unattractive girl or woman 10:hot dog 1
If you haven't studied semiotics(or even if you have!), you may be confused by an important distinction we will be making between the semiotics of an image and its grammar. Here is a relatively simple way to understand that distinction.
Table of Contents
Semiotics
Semiotics was formulated by the linguist Ferdinand de Saussure (let's leave aside for now its earlier formulation by C.S. Pierce). Saussure wanted to understand how meanings (signifieds) map to speech sounds (signifiers - e.g. words or lexemes). What struck him is that the signifiers are arbitrarily related to what they signify, and yet the signified is effectively impressed on the signifier (the material entity - an acoustic event or marks on a page). We experience the signifier as if it just were the signified. Indeed, it is virtually impossible not to hear the "meaning" of lexemes in your native language.
Saussure also noted something else of importance. Signifiers tend to form closed systems, i.e. there are only so many elements in a set of possible signifiers - a deck of cards, allowed sounds in a particular language. Thus the value of a particular signifier can often be defined by its position in a system - the ten of clubs in a standard deck of cards.
Given this framework, an obvious question to ask is whether the signifieds are also part of closed systems. Unfortunately, many people who pursue this question fail to understand the basic semiotic paradigm and therefore conflate two levels of analysis. The fact that the relationship between signifiers and signifieds in natural speech is arbitrary does not mean this relationship will be arbitrary in other semiotic systems. Futhermore, the fact that signifiers are part of a closed system does not mean that the signifieds must also belong to a structured system of meanings.
To understand the importance of this point, consider the following example. In the simple system of signifiers in the table on the left, I selected five vowel sounds from the phonological feature space shown on the right and placed each one between two consonants - "p" and "n." The signifieds for these five lexemes are the core entries each one generated in the online Marion Webster dictionary.
pan
pen
The word you've entered
isn't in the dictionary.
The first column in the table on the left illustrates the principle of positional variation for signifiers. The five possible signifiers in this simple system are differentiated purely on the basis of contrastive distinctions between the positions of vowels in the feature space on the right, e.g. pen / pin. The mapping of the signifieds to these five signifiers illustrates the principle of arbitrariness. There is no intrinsic or motivated reason other than purely contingent historical factors for any of the possible "meanings" to map to any of the signifiers. In fact, one of the possible signifiers - pon - has no signifieds. Moreover two of the signifiers are highly polysemic - i.e. map to many different entries in a lexicon - while others such as "pun" are less so.
If you look at the dictionary entries for the highly polysemic signifiers such as "pan," you will very quickly see that it is not possible to make generalizations about either the arbitrariness or the semantic structure of the signifieds. Consider, for example, just one of the entries for "pan":
1 a : a usually broad, shallow, and open container for domestic use (as for cooking ) b : any of various similar usually metal receptacles: as (1) : the hollow part of the lock in a firelock or flintlock gun that receives the priming (2) : either of the receptacles in a pair of scales (3) : a round shallow usually metal container for separating metal (as gold) from waste by washing c British : toilet 3b; also : bowl 3b d : steel drum
2 a (1) : a natural basin or depression in land (2) : a similar artificial basin (as for evaporating brine) b : a drifting fragment of the flat thin ice that forms in bays or along the shore.
There is an obvious motivated connection for most of the meanings in 1 and 2, namely what psychologists call affordance , i.e. "a quality of an object, or an environment, that allows an individual to perform an action." Even the signified in 2b can be taken as a semantic extension of the notion of a broad, shallow container.
Yet, even though many of these signifieds for "pan" have a deep motivated connection, they do not, either with respect to each other or within their own frame of reference, demonstrate the kind of positional variation found in the feature space for vowels. In other words, cooking pans, gun pans, and mining pans do not form a closed system. Moreover the domain of cooking utensils can be carved up many different ways by different cultures or at different periods of time (or by different manufacturers). Thus the contrastive distinction "pot / pan" at the level of signifieds is very much not the same as the distinction at the level of signifiers.
Semiotics is thus concerned with the relationship between signifiers and signifieds, a relationship that couples two very different things. The lesson here is that social selection on signifiers and signifieds also operates differently. Selection on a closed system of signifiers requires a remapping over a finite set of elements; selection on signifieds is highly contingent and does not necessarily involve a remapping across an entire feature space.
Grammar
A grammar is an account of how we map within the domain of signifiers or signifieds from an underlying representation (UR) to a surface representation (SR). In the tree diagram below, the German phrase "Wirtschafts und Währungs union" is the SR and the tree structure is the UR which maps the semantic properties in the set below the phrase to the SR.We can also construct a grammar of the signifiers from the semiotic example, that is a grammar of how vowels and consonants combine to form syllables. Below are two examples of such a grammar - one from a classical treatment of phonology and the other from a more recent neural network based account of phonology called Optimality Theoretic grammar.
classical phonology
syll-paramOT phonology
Paul Smolensky, Cognitive Science, Johns Hopkins Univ.Dog, a highly variable domestic mammal
What Kress is concerned with in Reading Images is the grammar of how signifiers in an image go together to make a representation. This level of analysis must be distinguished from any sort of semiotic "interpretation" we wish to give to the signifieds associated with the image.
non-transactional reaction - reacter
offer / social / detachment / equality
abstract coding orientation - sociocultural elites
symbolic attributive
demand / social / involvement / equality
naturalistic coding orientation - "just being themselves"
Dog
Function: noun Usage: often attributive Etymology: Middle English, from Old English docga Date: before 12th century
1 a : canid; especially : a highly variable domestic mammal (Canis familiaris) closely related to the gray wolf b : a male dog ; also : a male usually carnivorous mammal
2 a : a worthless or contemptible person b : fellow, chap <a lazy dog> <you lucky dog>
3 a : any of various usually simple mechanical devices for holding, gripping, or fastening that consist of a spike, bar, or hook b : andiron
4 : uncharacteristic or affected stylishness or dignity <put on the dog>
5 capitalized : either of the constellations Canis Major or Canis Minor
6 plural : feet
7 plural : ruin <going to the dogs>
8 : one inferior of its kind <the movie was a dog>: as a : an investment not worth its price b : an undesirable piece of merchandise
9 : an unattractive person; especially : an unattractive girl or woman
10 : hot dog 1